scieee Science in your language
[en] (orig)

Investigation of single point incremental forming parameters and forming limit curves prediction for heterostructured aluminum sheets

Author: Magalhães, Danielle Cristina Camilo,Montanari, Luciana,Elizalde Huitrón, Sergio Alberto,Cabrera Marrero, José M.,Benaque Rubert, José,Evangelhista, Sergio Henrique,Kliauga, Andrea Madeira
Year: 2025
DOI: 10.1016/j.matchar.2025.114993
Source: https://upcommons.upc.edu/bitstream/2117/430217/1/preproof%20Mater%20Charact.pdf
Jou nal P e-p oo
In es iga ion o single poin inc emen al o ming pa ame e s and
o ming limi cu es p edic ion o he e os uc u ed aluminum
shee s
Danielle C is ina Camilo Magalhães, Luciana Mon ana i, Se gio
Albe o Elizalde Hui on, Jose Ma ia Cab e a Ma e o, José
Benaque Rube , Se gio Hen ique E angelhis a, And ea Madei a
Kliauga
PII: S1044-5803(25)00282-7
DOI: h ps://doi.o g/10.1016/j.ma cha .2025.114993
Re e ence: MTL 114993
To appea in: Ma e ials Cha ac e iza ion
Recei ed da e: 3 Decembe 2024
Re ised da e: 28 Feb ua y 2025
Accep ed da e: 28 Ma ch 2025
Please ci e his a icle as: D.C.C. Magalhães, L. Mon ana i, S.A.E. Hui on, e al.,
In es iga ion o single poin inc emen al o ming pa ame e s and o ming limi cu es
p edic ion o he e os uc u ed aluminum shee s, Ma e ials Cha ac e iza ion (2024),
h ps://doi.o g/10.1016/j.ma cha .2025.114993
This is a PDF ile o an a icle ha has unde gone enhancemen s a e accep ance, such
as he addi ion o a co e page and me ada a, and o ma ing o eadabili y, bu i is
no ye he de ini i e e sion o eco d. This e sion will unde go addi ional copyedi ing,
ypese ing and e iew be o e i is published in i s inal o m, bu we a e p o iding his
e sion o gi e ea ly isibili y o he a icle. Please no e ha , du ing he p oduc ion
p ocess, e o s may be disco e ed which could a ec he con en , and all legal disclaime s
ha apply o he jou nal pe ain.
© 2025 Published by Else ie Inc.
Jou nal P e-p oo
Jou nal P e-p oo
In es iga ion o single poin inc emen al o ming pa ame e s
and o ming limi cu es p edic ion o he e os uc u ed
aluminum shee s
Danielle C is ina Camilo Magalhães1*, Luciana Mon ana i2, Se gio Albe o
Elizalde Hui on3, Jose Ma ia Cab e a Ma e o3,4, José Benaque Rube 5,
Se gio Hen ique E angelhis a5, And ea Madei a Kliauga1
1Depa men o Ma e ials Enginee ing, Fede al Uni e si y o São Ca los, Rod. Washing on Luís,
km 235, 13565-905, São Ca los, SP – B azil.
2Depa men o Mechanical Enginee ing, São Ca los School o Enginee ing - S a e Uni e si y o
São Paulo, A . T abalhado São-Ca lense, 400, 13566-590, São Ca los, SP – B azil.
3Depa amen o de Ciencia e Ingenie ía de Ma e iales, EEBE - Uni e si a Poli ècnica de
Ca alunya, c/Edua d Ma is any 10-14, 08019, Ba celona – Spain.
4Fundacio Cen e CIM, c/ Llo ens I A igas, 12, 08028, Ba celona-Spain.
5Depa men o Mechanical Enginee ing, Fede al Uni e si y o São Ca los, Rod. Washing on
Luís, km 235, 13565-905, São Ca los, SP – B azil.
*Co esponding au ho : dccamilo@u sca .b
ABSTRACT
This s udy explo es he o mabili y and ac u e beha io o Al-based
he e os uc u ed ma e ials (HM) ab ica ed using Accumula i e Roll-Bonding
(ARB) and assessed h ough Single Poin Inc emen al Fo ming (SPIF). The HM
shee s, consis ing o al e na ing AA1050 and AA7050 laye s, we e p ocessed a
p ehea ing empe a u es o 450 °C and 500 °C. De ailed cha ac e iza ion using
SEM, EBSD, and TEM e ealed a ia ions in g ain size, c ys allog aphic ex u e,
and p ecipi a e dis ibu ions be ween p ocessing empe a u es, in luencing he
s eng h, ha dness and duc ili y o HM shee s. Tensile es s showed ha shee s
p ocessed a 500 °C exhibi ed highe elonga ion and imp o ed duc ili y compa ed
o hose p ocessed a 450 °C, a ibu ed o changes in size and dis ibu ion o
p ecipi a es. Fo ming Limi Cu es (FLC) and SPIF expe imen s demons a ed
supe io o mabili y a 500 °C, wi h he HM shee s achie ing highe c i ical wall
angles and ac u e s ains compa ed o AA7050 laye s alone. The esul s
Jou nal P e-p oo
Jou nal P e-p oo
highligh ed he in luence o ARB o induce mic os uc u al ea u es, including
esidual s esses and shea s ain localiza ion, on he o ming beha io o HM
shee s. The s udy unde sco es he po en ial o op imizing p ocessing condi ions
and ma e ial composi ions o enhance he mechanical pe o mance and
manu ac u abili y o HM aluminum shee s.
Keywo ds: single poin inc emen al o ming; accumula i e oll-bonding;
he e os uc u ed shee s; o ming limi cu es; ex u e; aluminum alloys.
1. INTRODUCTION
He e os uc u ed ma e ials (HM) ep esen a no el class o ma e ials
cha ac e ized by he p esence o he e ogeneous zones wi h signi ican ly di e en
mechanical o physical p ope ies [1]. These ma e ials we e de eloped o
add ess challenges associa ed wi h he e y low duc ili y and oughness ound in
homogeneous nanos uc u ed o ul a ine-g ained ma e ials, while s ill
main aining high s eng h le els [2,3]. One ema kable ea u e o HM is ha hei
in eg a ed p ope ies o en su pass p edic ions made by he ule-o -mix u es,
which conside s ha he s eng h is gi en by he olume ic ac ion o he
he e ogeneous zones. In his con ex , se e al ypes o HM ha e been de eloped,
including mul ilaye ed s uc u es comp ising dissimila ma e ials, such as Al
alloys [4-6].
Fo ins ance, Al he e os uc u ed shee s ha e a ac ed signi ican
a en ion in ecen yea s due o hei combina ion o s eng h and duc ili y, making
hem candida es o he au omo i e and ae ospace indus ies [7, 8]. Recen
s udies ha e signi ican ly enhanced he unde s anding o he mechanical
esponse o HM, demons a ing hei po en ial o o e come he adi ional
s eng h-duc ili y ade-o and imp o e he o e all mechanical p ope ies o
mul ilaye ed shee s. Howe e , hese in es iga ions ha e p edominan ly
concen a ed on ensile p ope ies, such as yield s eng h, ul ima e ensile
s eng h, and elonga ion. This na ow ocus has le a c i ical gap in he
comp ehension o HM o mabili y, including hei beha io unde complex loading
condi ions, o ming limi s, and ac u e mechanisms, which is a majo
echnological aspec o manu ac u ing componen s.
Jou nal P e-p oo
Jou nal P e-p oo
Al hough nume ous s udies ha e been pe o med on single-laye and
bime allic shee s, esea ch on he o mabili y o mul ilaye ed shee me als
emains sca ce. The ea lies in es iga ions in o Fo ming Limi Cu es (FLC) o
bime allic shee s we e pionee ed by Semia in and Piehle [9], examining he
o mabili y o s ainless s eel/aluminum shee s. Thei wo k e ealed ha he
ini ia ion o necking and ac u e is highly dependen on he speci ic a angemen
o he laye s wi hin he shee . Building on his, Mo i and Ku imo o [10] u he
explo ed he o mabili y beha io o Al1100/SUS340 bilaye ed shee s. They
ound ha placing aluminum as he ou e laye signi ican ly enhances o mabili y
compa ed o al e na i e con igu a ions. Simila ly, Tseng e al. [11] s udied he
limi ing s ains o Al/Cu bime allic shee s du ing deep d awing p ocesses,
examining he e ec o hickness a ios be ween he laye s. Thei indings
highligh ed he de imen al impac o esidual s esses a he in e ace o he
laye s on o e all o mabili y. O he wo ks based on Al/Mg and Al/Cu we e also
pe o med in his opic [12, 13]. These s udies emphasize he complexi y o he
bime allic shee s beha io and he need o conside ac o s such as ma e ial
compa ibili y, laye hickness, spa ial dis ibu ion and esidual s esses when
p edic ing o mabili y in such sys ems.
Among a ious me hods o assessing shee o mabili y, he Single Poin
Inc emen al Fo ming (SPIF) echnique eme ges as a p omising app oach o
manu ac u ing complex geome ies o se e al indus ies [14-16]. Pa icula ly, he
SPIF p ocess has demons a ed a subs an ial enhancemen in he a ainable
s e ch and o mabili y o shee s compa ed o con en ional o ming o s amping
p ocesses when o mabili y/ ac u e cu es a e analyzed [17]. Consequen ly, i
becomes easible o p edic FLC using da a acqui ed h ough SPIF
expe imen a ion. In his con ex , SPIF can be applied o in es iga e he o mabili y
o se e al alloys, and pa icula ly Al alloys. The esul s ha e demons a ed ha
he o mabili y o Al alloys unde di e en p ocessing and es ing condi ions is
highe using SPIF han he hemisphe ical dome es [18-22].
Rega ding HM, only a ew s udies ha e used SPIF o analyze
de o ma ion and ac u e beha io . Some s udies p edic ed he necking and
ac u e s ain du ing SPIF using bi- o i-me allic shee s p oduced by explosi e
welding [23], o by Accumula i e Roll-Bonding (ARB) [24, 26]. In gene al,
de o ma ion and ac u e beha io s a e s ongly a ec ed by he laye s'
Jou nal P e-p oo
Jou nal P e-p oo
a angemen . In addi ion, he o ming angle may also be a ec ed by he laye in
con ac wi h he ool. Kuma e al. [27] in es iga ed SPIF o Al-Cu bime allic
shee s, emphasizing he ole o p ocess pa ame e s in achie ing high o mabili y
and su ace quali y wi hou delamina ion. Simila ly, Li e al. [28] explo ed he
o mabili y and ac u e beha io o mul ilaye ed shee s Al/s eel and Al/Ti shee s,
highligh ing he impo ance o laye hickness and oolpa h s a egies in
op imizing o ming limi s and educing ac u e. These s udies unde sco e SPIF's
capabili y o o m complex geome ies in mul ilaye ed sys ems while main aining
s uc u al in eg i y, hough u he esea ch is needed o e ine p ocess
pa ame e s o di e se ma e ial combina ions.
Despi e e y li le esea ch abou FLC and/o SPIF in HM, he
de elopmen o o mabili y cu es enables he p edic ion o c i ical p ocess limi s
and acili a es he design o e icien manu ac u ing p ocesses ailo ed o speci ic
applica ions o his new class o ma e ials. The cu en li e a u e highligh s he
impo ance o ma e ial p ope ies and geome ic ac o s, such as laye hickness,
in de e mining he o mabili y and pe o mance o HM shee s. Howe e , a
signi ican gap emains in he li e a u e, as he a o emen ioned s udies p ima ily
ocus on laye s wi h hicknesses in he millime e ange, lea ing he beha io o
mul ilaye sys ems wi h mic ome e -scale laye s la gely unexplo ed.
Unde s anding he o mabili y, ac u e mechanisms, and o e all pe o mance o
such hin mul ilaye con igu a ions is c i ical o indus ial applica ions. Fu he
esea ch is needed o b idge his gap and p o ide insigh s in o he unique
challenges and oppo uni ies p esen ed by mic ome e -scale mul ilaye ed HM
shee s.
In his s udy, we in es iga ed he o mabili y and ac u e beha io o HM
Al shee s composed o al e na ing AA1050 and AA7050 laye s wi h a e age
hickness below 25 μm, ab ica ed using ARB and assessed h ough SPIF. HM
shee s we e p ocessed by ARB a p ehea ing empe a u es o 450 °C and 500
°C, and hei mac o- and mic os uc u al ea u es, including laye s’ hickness,
mean g ain size, c ys allog aphic ex u e, and p ecipi a e dis ibu ions, we e
cha ac e ized. In addi ion, SPIF expe imen s we e pe o med o assess
o mabili y, measu ing c i ical wall angles and ac u e s ains. Thus, FLC we e
p edic ed using SPIF da a. These esul s we e co ela ed wi h mic os uc u e
ea u es and mechanical p ope ies o p o ide deepe insigh s in o he

Jou nal P e-p oo
Jou nal P e-p oo
de o ma ion mechanisms, o mabili y and ac u e beha io o HM shee s. The
aim o his s udy lies in i s po en ial o ad ance he adop ion o mul ilaye ed
aluminum shee s in di e se indus ial sec o s, o e ing ligh weigh , high-s eng h
solu ions o complex componen ab ica ion. By elucida ing he in ica e in e play
be ween ma e ial p ope ies, p ocess pa ame e s, and o ming beha io , his
esea ch con ibu es o he op imiza ion o inc emen al o ming p ocesses.
2. EXPERIMENTAL PROCEDURE
In his in es iga ion, HM shee s we e ab ica ed using ARB p ocessing,
which p o ides a good bonding s eng h be ween laye s. Comme cial shee s o
AA1050 (0.001 w .% Zn, 0.003 w .%Mg, 0.013 w .% Cu, 0.07 w .% Si, 0.12 w .%
Fe, and Al balance) and AA7050 (6.45 w .% Zn, 2.38 w .% Mg, 2.48 w .% Cu,
0.08 w .% Si, 0.09 w .% Fe, 0.09 w .% o he s, and Al balance) we e employed in
his in es iga ion. The pla es we e cu down in o specimens wi h a wid h o 50.0
mm and a leng h o 80.0 mm, and hen he ini ial shee s we e p oduced o
ARB.Fo his wo k, AA1050 and AA7050 shee s we e cleaned and s acked
al e na ely, and hen p ocessed up o six ARB cycles wi h p ehea ing o 5 min a
450 °C and 500 °C, wi h a inal pass wi h 50% hickness educ ion, which esul ed
in 1 mm shee s.
The cha ac e iza ion o he laye ed mac os uc u e o he HM
AA1050/AA7050 was pe o med by means o scanning elec on mic oscopy
(SEM) in a FEI-Inspec S50 mic oscope. All analyses we e aken on he RD-ND
plane ( olling di ec ion – no mal di ec ion). To measu e he indi idual hickness
a ia ion in each laye , a leas six s aigh lines we e ma ked on he
pho omic og aphs a he same magni ica ion, co e ing app oxima ely en laye s
in he same ame. The hickness o indi idual laye s and he mean s anda d
de ia ion alues we e es ima ed o bo h s udied condi ions. In addi ion, elec on
backsca e di ac ion (EBSD) analysis was pe o med in hese samples o
measu e he a e age g ain size and g ain bounda y cha ac e as well as o
de e mine he meso ex u e in o he laye s. Fo his pu pose, a Tescan MIRA3
SEM mic oscope was used, wi h ope a ing ol age o 25 kV and s ep size o 0.10
μm on he RD-ND plane. P io o EBSD, he samples we e mechanically polished
using diamond suspension, ollowed by elec ochemical polishing a -30 °C using
Jou nal P e-p oo
Jou nal P e-p oo
a solu ion wi h 14% o dis illed wa e , 6% o pe chlo ic acid and 80% o e hanol.
T ansmission Elec on Mic oscopy (TEM) samples we e p epa ed om he
mul ilaye ed composi e shee s using a winje elec o polishe wi h a solu ion o
60 mL o pe chlo ic acid, 140 mL o dis illed wa e and 800 mL o e hanol, wi h
an ope a ing ol age o 35 V. Obse a ion was pe o med in an FEI TECNAI
G²F20 TEM mic oscope ope a ed a 200 kV.
The c ys allog aphic ex u e was also measu ed by X- ay di ac ion a
he LNNano Synch o on Na ional Labo a o y in Campinas/B azil on a
PhilipsX’Pe MPD di ac ome e using Co-Kα adia ion wi h he Schulz me hod
o ob ain incomple e pole igu es (PF), backg ound and de ocusing cu es o
u he co ec ion. The o ien a ion dis ibu ion unc ions (ODFs) we e calcula ed
om he pole igu es o he {111}, {200} and {220} measu ed a he sub-su ace
o he shee s a e emo ing ¼ o he hickness by g inding and elec ochemical
polishing.
To comple e he ini ial cha ac e iza ion o he laye s, Vicke s
mic oha dness (HV) and ensile es s we e pe o med. Fo Vicke s
mic oha dness, a Fu u eTech ha dness es e was used wi h a load o 100 g
applied on he polished samples’ su ace (RD-ND plane) du ing 15 s. Mo e han
en inden a ions we e measu ed in each laye o es ablish he a e age ha dness.
P io o o mabili y es s, he shee s we e es ed in uniaxial ensile es s.
Subsize ensile specimens we e machined wi h dimensions o 3.0 × 2.0 × 7.0
mm in he gauge leng h, ollowing he RD. They we e es ed in an Ins on 5500
machine, a oom empe a u e and unde an ini ial s ain a e o 1.0 × 10-3 s-1. A
leas h ee specimens we e es ed o each condi ion and he p esen ed esul s
a e a e age alues. A e ensile es s, he ac u ed su aces we e analyzed by
SEM (FEI-Inspec S50 mic oscope).
To es ima e he o mabili y o he shee s, ini ially Nakazima es s we e
ca ied ou using i e di e en specimen geome ies o p oduce di e en plane
s ess condi ions. Th ee specimens o each geome y we e used o es ima e he
o mabili y cu es. These es s used a hemisphe ical punch wi h a 50 mm
diame e , ins alled in a Zwick BUP200 machine, wi h a punch speed o 1 mm/s
un il specimen ac u e. To de e mine he mino and majo s ain alues, an in
si u digi al image co ela ion (DIC) sys em was used du ing Nakazima es s, in
which specimens’ su aces we e p e iously pain ed wi h a whi e coa wi h black
Jou nal P e-p oo
Jou nal P e-p oo
speckle. Thus, he limi s ains we e es ima ed by he posi ion-dependen me hod
using he DIC sys em. Fo hese es s, lub ican oil and Te lon laye s we e used
o minimize ic ion. A e es s, he ac u ed su aces we e analyzed by SEM
(FEI-Inspec S50 mic oscope).
The SPIF expe imen s u ilized blanks measu ing 55 mm × 55 mm × 1
mm, which we e p epa ed by elec ochemically e ching g ids o 3 mm diame e
ci cles. This p epa a ion acili a ed he measu emen o in-plane s ains om
de o med ellipses. The expe imen al se up was housed wi hin he CNC
machining cen e (ROMI D600). The ini ial o ming diame e was 42 mm, and he
ool had a ip diame e o 4 mm. The ool pa h s a egy in ol ed ini ia ing wi h Ψi
and inc emen ally inc easing he d awing angles by ΔΨ = 5° un il eaching ei he
ac u e o he ool ip diame e limi o he benchma k conical shape. Di e en
ini ial Ψi angles we e es ed a ying om 20° o 55°.
Figu e 1. Inc emen al shee o ming. (a) Schema ic ep esen a ion o he ooling
se -up. (b) D awing angle (Ψ) and dep h (h).
The o ming ool had a hemisphe ical ip wi h 4 mm diame e and was
made om VC131 ool s eel, ha dened and empe ed o 60 HRC. The es s we e
pe o med wi h helical ool pa hs wi h a eed a e o 0.25 mm/ e (downwa d eed)
equal o a linea speed o 125 mm/min. The o a ion o he o ming ool was 1750
pm.
3. RESULTS
3.1 Ini ial cha ac e iza ion o he HM AA1050/AA7050 shee s
Fig. 2 shows he SEM images o HM AA1050/AA7050 shee s p oduced
a 450 °C and 500 °C, and hei espec i e a e age hickness and Vicke s
mic oha dness. In Figs. 2(a) and 2(b), i can be obse ed ha he laye s emained
app oxima ely s aigh and con inuous, wi h sligh wa iness. The obse ed
wa iness is due o di e ences in low s ess be ween he ma e ials, leading o
de o ma ion he e ogenei ies and he o ma ion o shea bands ha cu h ough
Jou nal P e-p oo
Jou nal P e-p oo
se e al laye s. Ano he ele an poin ha can be obse ed in hese mic og aphs
is he excellen adhesion be ween he laye s. In pa icula , he las laye
gene a ed in he inal cycle o ARB is well bonded due o he addi ional pass wi h
50% hickness educ ion. The s ain in he di usion p ocess, known as
de o ma ion-induced in e di usion, enhances elemen di usion in o he in e ace
zone h ough mechanisms such as acancy gene a ion, pipe di usion along
disloca ions, and s ess-induced a omic displacemen s. In he ARB p ocess,
comp ession and shea de o ma ion p omo e in e di usion, con ibu ing o
in e ace o ma ion and bonding s eng h, whe e mechanical bonding occu s
apidly du ing ARB, while di usion bonding du ing annealing plays a c i ical ole
in o ming AA1050/AA7050 in e aces [29].
Rega ding he laye hickness (Fig. 2(c)), i is obse ed ha he e is a
dispe sion o alues, wi h laye s a ying in hickness. Fo he case o shee s
p ocessed a 450 °C, he a e age hickness o AA1050 laye s is abou 23 ± 7 μm,
and o AA7050 laye s is 17 ± 6 μm. Fo samples p oduced wi h p ehea ing a
500 °C, he a e age hickness o AA1050 and AA7050 laye s was 24 ± 7 μm and
18 ± 8 μm, espec i ely.
Figu e 2. Mac os uc u es ob ained o HM AA1050/AA7050 shee s p oduced by
ARB a : (a) 450 °C e (b) 500 °C. (c) Thickness a ia ion o he in e nal laye s. (d)
A e age Vicke s ha dness measu emen s in indi idual laye s.
Rega ding he ha dness e olu ion o indi idual laye s, Fig. 2(d) shows
ha he AA1050 Al laye s exhibi simila ha dness le els a bo h p e-hea ing
empe a u es, a ound 50 ± 5 HV0.1. Ini ially, he ha dness le el o AA1050 Al
shee s was 20 ± 1 HV0.1. This sugges s ha he inc ease in ha dness is due o
ei he esidual s ain om he eco e ed subs uc u e and/o solu e s eng hening
om di usion (o igina ing om AA7050 Al laye s) du ing p ocessing. Con e sely,
he AA7050 laye s displayed a ha dness le el lowe han hei ini ial condi ion
(150 ± 7 HV0.1) a bo h p ocessing empe a u es. The p ima y ac o con ibu ing
o he ha dness o his alloy is he p ecipi a es’ size and dis ibu ion. Shee s
p ocessed a 450 °C eached a lowe ha dness alue, app oxima ely 121 ± 4
HV0.1. When p ocessed a 500 °C, he highes ha dness alue o abou 137 ± 5
Jou nal P e-p oo
Jou nal P e-p oo
shee s, wi h hei dominan shea componen s, exhibi ed highe esis ance o
de o ma ion, while he AA1050 laye s displayed g ea e o mabili y, e en wi h a
mo e p onounced olling ex u e. This in e ac ion leads o aniso opic beha io ,
highligh ing he need o u he explo a ion o s ess dis ibu ion and s uc u al
in eg i y a e o ming.
In e es ingly, HM shee s achie ed highe dep hs o all d awing angles
han AA7050 alloy alone. In addi ion, as FLC esul s indica ed, he shee s
p ocessed a p ehea ing empe a u e o 500 °C showed be e o mabili y han
ha a 450 °C. Fig. 11 p esen s pho og aphs o he specimens a e SPIF
expe imen s wi h 𝜂i = 25o, showing ha HM shee s wi h an in e media e beha io
be ween AA1050 and AA7050 sepa a ely. The ac u e ini ia ed a he ou e
su ace and hen p opaga ed owa ds he inne su ace. Because o he ini ial
geome y o he expe imen , he ac u e o he AA1050 shee was only obse ed
a ini ial wall angles highe han 35°. The ac u e was duc ile wi h he p esence
o necking and educ ion o he wall hickness. The c i ical wall angle achie ed
alues be ween 82° and 86°. The ac u e in he AA7050 shee was cha ac e ized
by he absence o necking, and a much smalle wall angle o 32° o 36°. The
ac u e o he HM s uc u es was cha ac e ized by delamina ion and he c i ical
wall angles inc eased o 39° - 44° and 39° - 50° o he p ocess empe a u es o
450 °C and 500 °C, espec i ely.
Table 3. Expe imen al esul s o es s p oduced by SPIF. Ini ial wall angle (Ψi)
and inal dep h and inal wall angle (Ψ ). No ac u e ◯; ac u e ⚫.
Ma e ial
Ψi (º)
Dep h
(mm)
Ψ (º)
F ac u e
AA1050
20
10.25
40.00
◯
25
13.50
51.50
◯
30
19.50
58.50
◯
35
27.25
67.00
⚫
40
25.00
85.00
⚫

Jou nal P e-p oo
Jou nal P e-p oo
Figu e 11. Rep esen a i e specimens’ pho og aphs a e SPIF expe imen s
ini ia ed wi h 𝜂i = 25°: (a) AA1050 shee - no up u e; (b) AA7050 shee - up u e
a 𝜂 = 36o; (c) HM shee p oduced a 450 °C - up u e a 𝜂 = 44°; (d) HM shee
p oduced a 500 °C - up u e a 𝜂 = 46°.
In SPIF he wall hickness diminishes wi h he inc emen o he dep h and
wall angle. Acco ding o Houssain e al. [33] he wall hickness can be es ima ed
by he geome ical condi ions o he expe imen , gi en by Eq (1):
𝑡=𝑡0𝑡𝑡𝑡 𝑡𝑡 (1)
Whe e 0 is he ini ial wall hickness. Fo each ma e ial, he e is a hinning
limi , and he es ima ed s ain in he wall di ec ion is shown in Fig. 12. The c i ical
s ain o he AA1050 alloy was -0.65 and o he AA7050 -0.17. The HM ma e ials
wi h a olume ic ac ion o 0.50 o each alloy yielded -0.23 and -0.30 o HM 450
°C and HM 500 °C, espec i ely.
Figu e 12. Rup u e s ains h ough he wall di ec ion (εwall) as a unc ion o he
ini ial wall angle (ψi).
45
22.75
83.50
⚫
50
20.25
82.50
⚫
55
14.75
86.00
⚫
AA7050
20
6.50
32.50
⚫
25
5.75
36.00
⚫
30
5.75
36.00
⚫
HM shee - 450 °C
20
10.00
39.50
⚫
25
11.25
44.00
⚫
30
6.50
42.50
⚫
HM shee - 500 °C
20
10.00
39.50
⚫
25
10.75
46.00
⚫
30
10.00
49.00
⚫
Jou nal P e-p oo
Jou nal P e-p oo
The esul s om he SPIF expe imen s on HM AA1050/AA7050 shee s
p o ided aluable da a o p edic ing FLCs. These cu es highligh he ma e ial's
capaci y o unde go plas ic de o ma ion unde inc emen al o ming. Compa isons
we e made wi h FLCs ob ained om con en ional Nakazima es s, as shown in
Fig. 13 and wi h da a published in he li e a u e o he AA1050 [34] and AA7050
alloys [35].
Figu e 13. Compa ison o FLC da a om SPIF expe imen s and FLC esul s o :
(a) AA1050, (b) AA7050, (c) 450 °C and (d) 500 °C. Fo e e ence, see Fig. 9:
dashed lines e e o FLC cu es o mul ilaye ed ma e ials p ocessed a 450 °C
and 500 °C.
4. DISCUSSION
FLC cu es p edic he occu ence o necking in con en ional shee
o ming and a e used as a p ac ical limi o duc ile ac u e [36]. Howe e , he
s ain o ac u e p esen s di e en ailu e mechanisms. The FLC canno be used
as a limi ing c i e ion o SPIF, and he F ac u e Fo ming Limi line (FFL) is
gene ally u ilized in his case [37].
The Hill c i e ion o necking ailu e [38] p edic s ha he poin o ailu e
depends on he ha dening exponen n o he Hollomon’s ha dening law (Eq. 2)
ha co ela es e ec i e s ess (𝜎𝑡𝑡𝑡), e ec i e s ain (𝜀𝑡𝑡𝑡) and a s eng h
coe icien (K):
𝜎𝑡𝑡𝑡 =𝑡𝑡𝑡𝑡𝑡 𝑡 (2)
Localized necking happens a a condi ion ε1 = n o plane s ain ensile
s aining, as shown in Table 2, and di use necking occu s a ε1 = ε2 = n o equi-
biaxial ensile s aining. Because he onse o necking and oid nuclea ion occu s
di e en ly depending on he s ess s a e, wo se s o equa ions a e no mally
applied o p edic he FLC cu e: one o he le side (ε2 < 0), and ano he o he
igh side (ε2 > 0). Paul [36] p oposed he ollowing Eqs. (3) and (4):
Fo 𝛽=𝑡2
𝑡1<0: 𝜀1=𝑡𝑡𝑡0−𝑡2 (3)
Jou nal P e-p oo
Jou nal P e-p oo
Fo 𝛽=𝑡2
𝑡1>0: 𝜀1=(1+𝑡𝑡𝑡0)(1+𝑡2)𝑡−1 (4)
FLC0 and p a e i ing pa ame e s, ob ained by linea eg ession using
UTS, n, 𝑡1 om he ensile es , , and shee hickness , gi en by Eq. (5) and (6):
𝑡𝑡𝑡0= 7.702𝑡𝑡𝑡(−0.0122 𝑡𝑡𝑡)−0.1124 𝑡−0.690𝑡𝑡𝑡(−12.4187𝑡1)+
+0.01149𝑡 +0.0823 𝑡+0.3011 (5)
𝑡=1.0834 𝑡𝑡𝑡(−1.4114 𝑡𝑡𝑡0)−0.361 (6)
Fig. 14 compa es he p edic ed FLC using Equa ions (3-6) and da a
om Tables 1 and 2. The expe imen al da a is shi ed o he igh , which is a
cha ac e is ic o p e-s ained ma e ial. I was shown by he esidual s ain
analysis in Fig. 3 ha he AA1050 laye s ha e highe densi y o geome ically
necessa y disloca ions, and hus accumula e highe miso ien a ion wi hin he
g ains han in he AA7050 laye s. Due o he di e ence in YS alues o he wo
laye s, he di e en s ain le els lead o he buildup o shea componen s a he
laye bounda ies and o highe s ain le els in he so laye [31]. Kümmel e al.
[39] ha e also measu ed di e en esidual s ess le els in he aluminum MH
p oduced by ARB: comp essi e esidual s esses o m in he so laye s and
ensile esidual s esses in he ha d laye s. The esidual s esses inc ease wi h
mo e p onounced di e ences in YS o he cons i uen ma e ials. A he
p ocessing empe a u es, a highe ha dening s ess di e ence is achie ed a 450
°C in compa ison wi h 500 °C. The e o e, highe esidual s ess g adien s a e
expec ed in he HM AA1050/AA7050 shee p ocesses a 450 °C.
Figu e 14. Compa ison be ween p edic ed FLC cu es using equa ions
sugges ed by Paul [35] and he expe imen al da a.
The same esul s o he FLC may be p esen ed in he o m o e ec i e
s ain e sus s ess iaxiali y η, de ined as he a io be ween he mean p incipal
Jou nal P e-p oo
Jou nal P e-p oo
s ess and he on Mises e ec i e s ess (𝜂=𝑡𝑡
𝑡𝑡𝑡𝑡). This can be calcula ed om
he main s ain alues by Eq. (7) [40]:
𝜂=√3
3
1+𝑡
√1+𝑡+𝑡2 (7)
Whe e 𝜂 is he a io o he mino o majo in-plane s ains (𝜂2/𝜂1).
The e ec i e s ain acco ding o he Hill c i e ion [38] is calcula ed by Eq.
(8):
𝜀𝑡=1+𝑡𝑡
√1+2𝑡𝑡√𝑡1
2+𝑡2
2+2𝑡𝑡
(1+𝑡𝑡)𝑡1𝑡2 (8)
This limi o he duc ile ac u e is shown on Fig. 15. The ac u e mode
in ensile plane s ess expe imen s leads o duc ile damage by oid nuclea ion–
oid g ow h– oid coalescence (in mode a e o high iaxiali ies). The ARB
bonding o he pai AA1050/AA7050 is cons i u ed o a duc ile, low s eng h
ma e ial and a high s eng h wi h low duc ili y ma e ial. The ac u e o he HM
shee in ension and in plane s ess is cha ac e ized by duc ile ac u e in he
AA1050 laye and agile ac u e in he AA7050 laye . The p incipal idea is ha
he duc ile phase should imp o e he o mabili y o he HM, bu c ack ini ia ion in
he AA7050 laye s should also be a oided. I can be obse ed in Figs. 8, 10 and
13 ha he a ea o duc ile ac u e inc eases a a highe annealing empe a u e.
I is shown in Fig. 4 ha a bo h empe a u es he T-phase (Al3ZnMg) and S-
phase (Al2CuMg) coalesce wi h he inc emen o ARB cycles and he olume ic
ac ion o p ecipi a es is highe in he ma e ials p oduced a 450 °C. The T-phase
is coa se , and he S-phase is pa ially dissol ed in he HM500 ma e ial. In
addi ion, because he sol us empe a u e o he η- o η′-phase (MgZn2) lays a
475 °C, he e will be dissolu ion and ep ecipi a ion a 500 °C and coalescence
o his phase a 450 °C. Thus, he imp o ed o mabili y o HM shee s p ocessed
a 500 °C can be a ibu ed o he e ined p ecipi a e dis ibu ion and la ge g ain
size when compa ed o he 450 °C p ocess, as shown in Figu es 2, 3, and 4. Fine
p ecipi a es ac as e ec i e ba ie s o disloca ion mo ion, enhancing s eng h,
while he la ge g ain size imp o es duc ili y. These mic os uc u al changes
Jou nal P e-p oo
Jou nal P e-p oo
esul in a unique balance o s eng h and duc ili y, which is c i ical o he
o mabili y o HM shee s unde SPIF.
Figu e 15. Compa ison o he duc ile ac u e limi in he iaxiali y s. necking
s ain diag am o he ma e ials o his s udy.
The main ad an ages o he SPIF p ocess a e he simplici y o he
p ocess con igu a ion and he ex ension o he s ain o ac u e obse ed in
duc ile ma e ials. Howe e , he s ess-s ain s a e is e y complex, and he
localiza ion and ac u e happen a highe equi alen s ains due o he na u e o
he de o ma ion.
Eykens e al. [41] and Malho a e al. [42] obse ed ha he supp ession
o he necking mechanism is due o bending plus shea pa allel o he su ace
caused by he ool ansla ion. Acco ding o hem, his esul s in a ac u e due o
an ou -o -plane shea . A e e alua ion o measu ed and simula ed s ains, hey
conclude ha a pu e shea mechanism is unlikely because o he bending
s i ness o he shee . The local bending o he shee a ound he ool causes a
g ea e plas ic s ain a he ou e side o he shee while he inne pa is subjec ed
o shea and hyd os a ic p essu e (comp ession in he con ac poin and he
me idional and ci cum e en ial di ec ions). The e is also a h ough hickness
shea g adien componen . The e o e, he obse ed s ains a he ou e su ace
a e no longe he p incipal s ain di ec ions and i is no possible o de e mine a
single s ess s a e ac oss he shee hickness. Fo example, Mi nia and Shamsa i
[43] calcula ed a s ess iaxiali y o -0.6 a he inne su ace and +0.6 a he ou e
su ace in he egion o con ac wi h he SPIF ool and a s ong a ia ion o
in ensi y as he ool p og esses.
Li e al. [44] concluded ha he e is a combina ion o shea ing, bending
and s e ching. In cone- o ming, he majo de o ma ion, acco ding o he au ho s,
is he s e ching pe pendicula o he o ming di ec ion. The majo shea s ain
lies in he o ming di ec ion and is maximum a he middle o he shee . They also
ound ha he in-plane shea s ain is no negligible.

Jou nal P e-p oo
Jou nal P e-p oo
In addi ion, he con ibu ion o bending o ac u e in he SPIF p ocess is
cha ac e ized by a c i ical wall angle (𝜂). Fo he AA1050 alloy he epo ed
c i ical angle o conical us um and conical shapes lay be ween 65- 85° [45, 46],
and he p esen esul s a app oxima ely 85o a e in acco dance wi h he li e a u e.
Hussain e al. [33] ha e also no ed ha he poin o wall angle in he ailu e
depends on he geome y o he shape o med and may exceed he es ima e
ob ained by cons an wall angle pa s by ∼4°. The es ima ed s ain in he di ec ion
o he wall was -0.65. The AA7050 alloy on he con a y shows e y low plas ici y,
and he ac u e occu s wi hou necking wi h an es ima ed s ain in he wall
di ec ion o -0.17, and a low c i ical wall angles. The HM shee ends o be he
same beha io , bu he s ain limi was ex ended wo old a 500 °C.
Ma ins [47] examines he ac u e modus and loci unde plane s ess
condi ions. Fo he AA1050 alloy, he SPIF p ocess is cha ac e ized by a mode I
ac u e mode wi h supp ession o necking and he e ec i e s ain is limi ed by a
cons an hough- hickness e ec i e s ain (as seen in Fig. 11) and by ac u e
oughness. The ac u e o ming limi (FFL) is a s aigh line alling om le o
igh in he p incipal s ain space. Fo he AA1050 alloy his limi yield is gi en by
Eq. (9) [48]:
𝜀𝑡1 =1.36 −0.7𝑡𝑡2 (9)
The ac u e is ela ed o oid g ow h, which is associa ed wi h he
ac u e o he in e me allic pa icles o diame e d embedded in he aluminum
ma ix and he s ess iaxiali y. In he case o plane s ess and a no mal plas ic
aniso opy 𝑡𝑡 (Eq. 10) [49]:
𝑡𝑡𝑡𝑡𝑡
𝑡=𝑡𝑡 (1
𝑡)=∫𝜎𝑡
𝑡𝑡𝑓𝑡
𝜀𝑒𝑓𝑓
0𝑡𝜀=1+𝑟𝑚
3(𝜀1𝑡+𝜀2𝑡) (10)
Madei a e al. [48] ha e measu ed s ain necessa y o p opaga e a c ack
in ension (mode I) o samples o he AA1050 alloy and compa ed he esul s
Jou nal P e-p oo
Jou nal P e-p oo
wi h he SPIF expe imen s wi h di e en pa / oll diame e s. When his a io was
la ge han 40, no necking was obse ed and he SPIF limi coincided wi h he
mode I ac u e limi . Bu , o alues o pa / oll diame e s smalle han 25 (10 in
he p esen wo k), ailu e was p eceded by necking. In his case, he onse o
ailu e is delayed by dynamic bending unde ension. The ac u e in bending is
hen con olled by he a io be ween he shee hickness and he adius o he
o ming ool. Fo a pa / ool diame e a io o 10, hey ob ained a c i ical e ec i e
s ain o ac u e o 0.77. Fig. 16 shows he iso-s ain locus o ailu e in plane
s ess (FLC), o mode I ac u e (FFL), and he locus o e ec i e s ain 𝜂e = 0.8
in compa ison wi h he majo and mino s ain alues ob ained in he p esen
esul s. In he p esen esul s, a ac u e locus o a mixed ac u e mode was
ob ained, bu i is no iced ha o smalle ini ial 𝜂 angle (35-40°) he s ain o
ac u e was educed, and he s ain slope 𝜂 shi s o he comp ession- ension
quad an , because in hese condi ions he bending componen is inc eased wi h
espec o he s e ching componen .
Figu e 16. F ac u e loci ob ained in he SPIF expe imen s o he AA1050 alloy
in his wo k compa ed wi h he FLC and FFL limi s ob ained by Madei a e al.
[47].
Because o he lowe plas ici y, he SPIF expe imen s o he AA7050
alloy, and he HM ma e ials we e pe o med wi h a low ini ial wall angle. Unde
his condi ion he bending componen is highe han he s e ching componen .
The s ain pa h and he ac u e locus occu ed in he comp ession - ension
quad an . The ac u e o he HM shee s ook place a he ou e su ace and wi h
sigh s o delamina ion as shown in Fig. 11.
Muhammad e al. [49, 50] analyzed he bendabili y o an age ha dened
AA6061 aluminum alloy and obse ed ha he ini ia ion and p opaga ion o
c acks we e associa ed wi h he o ma ion o shea bands and h ough- hickness
g oo ing along he ou e ensile edge a he ensile side o he shee . Shea bands
lead o disloca ion accumula ion a g ain bounda ies, and in he p esence o
seconda y phase pa icles, decohesion o he bounda ies ook place. Shea
bands o med a he su ace may p opaga e inwa ds h ough mul iple g ains,
spanning ac oss he en i e wid h o he specimen. E en ually, his in ense
Jou nal P e-p oo
Jou nal P e-p oo
shea ing p omo es he o ma ion o su ace c acks which p opaga e inwa ds in a
ansg anula manne h ough he shea ed egions. This implies ha a he ou e
su ace plas ic s ain localizes in he di ec ions o maximum shea s ess and in-
plane shea ac s as a mechanism o ailu e. Cladding wi h low alloyed aluminum
led o an inc ease o bendabili y because he c ack nuclea ion mechanism was
shi ed o he so e ma e ial. Thinne cladding laye s lead o ailu e a he cladding
in e ace. This ac u e mode is simila o he ones obse ed in he AA7050 and
HM ma e ials.
Sil a e al. [51] discussed he locus o ac u e in he 𝜂1-𝜂2 s ain diag am
in shea (mode II). The de ec spacing 1/d is ela ed o he shea s ain 𝜂 by Eq.
(11):
𝑡𝑡 (1
𝑡)=√1+𝑡2 (11)
And he onse o damage (𝑡𝑡𝑡𝑡𝑡
𝑡𝑡 ) will ake place a he maximum shea
s ain is gi en by Eq. (12):
𝑡𝑡𝑡𝑡𝑡
𝑡𝑡 =𝑡𝑡 (1
𝑡)=∫𝜏
𝜎𝑒𝑓𝑓
𝜀𝑒𝑓𝑓
0𝑡𝜀= 1(1+𝑡𝑡)
2(1+2𝑡𝑡)(𝑡1𝑡−𝜀2𝑡) (12)
The c i ical s ain alues o damage in plane shea (SFFL) a e in s aigh
lines ising om le o igh wi h slope +1 in he comp ession ension quad an .
Fig. 17 shows he onse o ac u e ob ained o he AA7050 and he HM shee s.
The damage o he AA7050 alloy and he HM alloys is simila ( he shea s ains
𝜂 yield be ween 0.13 and 0.17), which indica es ha he ac u e limi s a e
con olled by he AA7050 laye s. Addi ionally, he delamina ion obse ed in SPIF
expe imen s was mo e p onounced in samples p ocessed a 450 °C, whe e
in e acial bonding was weake , as shown in Fig. 11. This phenomenon
nega i ely impac s he o e all pe o mance o he ma e ial by educing load
ans e be ween laye s and p omo ing c ack p opaga ion. To mi iga e
delamina ion, u u e wo k will ocus on op imizing ARB p ocessing condi ions and
explo ing he use o in e media e laye s o enhance in e acial bonding. Fo hese
ma e ials a simila s a egy as in o bending applica ions [50] using a hicke laye
o he AA1050 alloy a he shee su ace ei he by cladding o ARB may imp o e
he ou pu in he SPIF p ocess. Due o he highe shea componen s in his
Jou nal P e-p oo
Jou nal P e-p oo
p ocess, he adhesion be ween laye s is a c i ical pa ame e , much mo e han in
con en ional o ming, as shown by he FLC cu es (see Fig. 16).
Figu e 17. F ac u e loci ob ained in he SPIF expe imen s o he AA7050,
HM450 and HM500 ma e ials.
A deepe unde s anding o he delamina ion p ocess necessi a es u he
in es iga ion. One ac o ha enhances adhesion du ing ARB a ele a ed
empe a u es (450 and 500 °C) is he educ ion in he low s ess di e ence
be ween he laye s [31]. A oom empe a u e, howe e , he dispa i y in he
ha dening beha io o he AA1050 and AA7050 laye s becomes mo e
p onounced, leading o highe in e acial shea s ains ha may con ibu e o
ailu e. Fu u e esea ch will ocus on employing ad anced cha ac e iza ion
echniques, nume ical simula ions, and s a egies o mi iga e delamina ion,
he eby u he op imizing he pe o mance o HM shee s in he SPIF p ocess.
5. CONCLUSIONS
In his wo k, he po en ial o adop ing mul ilaye ed aluminum shee s in
o ming p ocesses was analyzed. A laye ed s uc u e consis ing o equal pa s o
AA1050 and AA7050 alloys, wi h a mean hickness below 25 μm, was p oduced
by ARB a p ocess empe a u es o 450 °C and 500 °C. The s udy highligh s he
c i ical ole o esidual s esses, pa icula ly hose a ising om he yield s eng h
di e ences be ween he laye s, in in luencing o mabili y and ac u e beha io .
The key indings o his wo k a e summa ized as ollows:
i) The AA1050 alloy exhibi s duc ile ac u e, while he AA7050 alloy
shows b i le ac u e a oom empe a u e, wi h he HM shee s
displaying in e media e beha io and imp o ed duc ili y and o mabili y
when p oduced by ARB wi h a p ehea ing empe a u e o 500 °C. In
HM, he AA1050 laye s a e cha ac e ized by a s ong olling ex u e
(e.g., B ass and S componen s), exhibi ed highe duc ili y and
o mabili y, con ibu ing o he o e all s e chabili y o he HM shee s.
In con as , he AA7050 laye s, wi h a highe ac ion o shea ex u e
Jou nal P e-p oo
Jou nal P e-p oo
Decla a ion o in e es s
☒ The au ho s decla e ha hey ha e no known compe ing inancial in e es s o
pe sonal ela ionships ha could ha e appea ed o in luence he wo k epo ed in his
pape .
☐ The au ho s decla e he ollowing inancial in e es s/pe sonal ela ionships which
may be conside ed as po en ial compe ing in e es s:

Jou nal P e-p oo
Jou nal P e-p oo
HIGHLIGHTS
 SPIF o mabili y o he e os uc u ed aluminum shee s was
sys ema ically explo ed.
 ARB a 450 °C and 500 °C e eals key mic os uc u al in luences on
o mabili y.
 Enhanced duc ili y and c i ical wall angles we e achie ed a e ARB a
500 °C.
 S a egies o acili a e he SPIF o he e ogeneous s uc u es a e
sugges ed.
52.